// deep dives

Long-form

  1. VLA-DSS: a Vision-Language-Action model on the efficiency frontier

    How a 28.9M-parameter, RGB-only robot policy holds its own against models several times its size, by trading learned visual capacity for mathematically-grounded structure: wavelet scattering, a frozen DINOv3 backbone, and a Fourier Neural Operator action head.

    • PyTorch
    • Vision-Language-Action
    • Robot learning
  2. The Axiom Journey: Engineering a High-Performance Engine

    A technical dive into the core architecture, sprite batching, and Entity-Component Systems (ECS) of the Axiom 2D game engine.

    • C++
    • OpenGL
    • Game Engine
// notes

Paper notes

Close reads and worked-through math, the things I wanted to remember exactly, kept by field as the work shifts.

// the reading list 30 papers

Vision-language-action & robot learning 9

Algorithms & systems 1